PSYCHOPATHY AND DECEPTION: DO PSYCHOPATHIC

PERSONALITY TRAITS MODERATE THE ABILITY

TO AVOID DETECTION OF DISSIMULATION?

by

BRANDEE E. MARION

MARTIN SELLBOM, COMMITTEE CHAIR

RANDALL T. SALEKIN

REBECCA J. HOWELL

A THESIS

Submitted in partial fulfillment of the requirements

for the degree of Master of Arts

in the Department of Psychology

in the Graduate School of

The University of Alabama

TUSCALOOSA, ALABAMA

2011

Copyright Brandee E. Marion 2011

ALL RIGHTS RESERVED

ii

ABSTRACT

Individuals high on psychopathic personality traits (e.g., callousness, impulsivity,

grandiosity) are likely to commit a high degree of criminal infractions, especially those of a

violent nature. When undergoing a forensic evaluation, these individuals may be motivated to

present themselves in an overly positive or negative light in order to obtain a favorable verdict or

sentence. Many personality inventories, such as the Minnesota Multiphasic Personality Inventory

– 2 – Restructured Form (MMPI-2-RF; Ben-Porath & Tellgen, 2008), contain validity scales

which are designed to detect such response bias. Study 1 of the current investigation sought to

determine whether psychopathy moderates the utility of the MMPI-2-RF‟s over-reporting (F-r,

Fp-r) and under-reporting (L-r, and K-r) validity scales in differentiating between individuals

asked to feign good, feign bad, or respond honestly. Study 2 replicated the over-reporting (F-r

and Fp-r) analyses in a forensic pre-trial sample, in which individuals were classified as

malingering or not malingering using the Structured Interview of Reported Symptoms (SIRS;

Rogers et al., 1992). Combined results for the over-reporting analyses indicated that psychopathy

did not moderate the utility of the MMPI-2-RF‟s validity scales in differentiating between honest

responders and those feigning symptoms of psychopathology. The under-reporting analyses

indicated no moderating effects for L-r; however, the “meanness” factor of psychopathy

moderated the utility of K-r in detecting those feigning psychological adjustment, such that K-r

was better able to detect individuals high on rather than low on psychopathy when under-

reporting. Implications of these results, as well as future directions, are discussed.

iii

DEDICATION

This thesis is dedicated to my mentor, Martin Sellbom, PhD., who has supported me

through every stage of this project. I am deeply appreciative of his encouragement, patience, and

understanding throughout this strenuous process. His never-ending dedication to intellectual

pursuits has inspired me to strive for excellence in my own life.

iv

LIST OF ABBREVIATIONS AND SYMBOLS

d Cohen‟s d effect size estimate

df Degrees of freedom

F Fisher‟s F ratio

M Mean

p Probability

R2 Coefficient of determination

T Standardized T-score

2

Pearson‟s chi-square statistic

< Less than

= Equal to

% Percentage of variance explained

v

ACKNOWLEDGMENTS

I am extremely thankful to the faculty members and colleagues who have provided

assistance with this project. Above all, I am grateful to my mentor, Martin Sellbom, for his

constant support, guidance, and hard work over the past two years. I especially appreciate his

willingness to share knowledge and provide detailed feedback, which have helped me grow as a

researcher. I would also like to express my sincere thanks to my other committee members, Drs.

Randall Salekin and Rebecca Howell, for their insightful feedback and valuable contributions to

this project. My graduate student colleagues have provided both an academic sounding board and

personal support, for which I am grateful. I thank Drs. Thomas Kucharski and Scott Duncan for

allowing me to use the forensic pre-trial sample for this project. I would also like to acknowledge

the undergraduate research assistants who assisted with data collection for this project: Tina

Wall, Naomy Velazquez, Melanie Childers, Ellen Super, Haley Barnett, Taylor Watts, and

Samantha Price.

My thanks also extend to my family and friends; without them, this project would not

have been possible. Thank you to Edgar Marion, Deborah Marion, and Nicholas Marion for

helping me remember what things are important in life. Finally, thank you to Sarah Malcom,

Michael Ricciardella, and Kathryn Lowe, for never-ending encouragement and humor.

vi

CONTENTS

ABSTRACT ................................................................................................ ii

DEDICATION ........................................................................................... iii

LIST OF ABBREVIATIONS AND SYMBOLS ...................................... iv

ACKNOWLEDGMENTS ...........................................................................v

LIST OF TABLES .................................................................................... vii

LIST OF FIGURES ................................................................................. viii

1. INTRODUCTION ...................................................................................1

2. STUDY 1 ...............................................................................................13

a. Methods ..................................................................................................13

b. Results ....................................................................................................19

3. STUDY 2 ...............................................................................................28

a. Methods ..................................................................................................28

b. Results ....................................................................................................30

4. GENERAL DISCUSSION ....................................................................34

REFERENCES ..........................................................................................42

APPENDICES ...........................................................................................49

vii

LIST OF TABLES

1. Eigenvalues and Factor Loadings for Exploratory Factor Analysis of

Psychopathy Scales ....................................................................................20

2. Means, Standard Deviations, F-tests, and Effect Sizes for Honest,

Under-reporting, and Over-reporting Groups ............................................22

3. Hierarchical Logistic Regression Analyses Predicting Over-reporting

vs. Not Over-reporting, or Under-reporting vs. Not Under-reporting,

Using Validity Scale, Psychopathy, and Psychopathy X Validity Scale. ..23

4. Means, Standard Deviations, F-tests, and Effect Sizes for Honest and

Over-reporting Groups ...............................................................................30

5. Hierarchical Logistic Regression Analyses Predicting Over-reporting

vs. Not Over-reporting Using Validity Scale, Psychopathy, and Psychopathy

X Validity Scale .........................................................................................31

viii

LIST OF FIGURES

1. Interaction Between Meanness and K-r Scores in Predicting Honest vs.

Feign Good Groups ....................................................................................26

1

Introduction

The psychopathic personality is characterized by an array of maladaptive attributes and

behaviors (Hare, 1996). These are usually classified as affective-interpersonal features (i.e.,

callousness, superficiality, and lack of remorse or empathy) and antisocial behavior styles (i.e.,

sensation seeking, impulsivity, and lack of responsibility; Hare, 1996). On the surface,

psychopaths often appear normal or even charming, while in reality they tend to be guiltless,

with no regard for the feelings or welfare of other people (Hare, 1996). Coupled with the

impulsivity and thrill seeking components, the psychopath‟s lack of concern for societal rules

increases the likelihood of criminal behavior (Hare, 1996; Hart & Hare, 1997). Psychopaths are

more likely to engage in criminal behavior at a young age, and are responsible for a large

proportion of violent crime and violent recidivism (Salekin, Rogers, & Sewell, 1996; Walters,

2003). Once incarcerated, psychopaths commit a large number of institutional infractions (Hart,

1998; Widiger, 2006).

Because psychopaths are often involved in the criminal justice system, they frequently

undergo forensic psychological evaluations (Hare, 2003). These individuals pose significant

threats to society; thus, it is vital that psychologists accurately evaluate these individuals (Edens,

2006). Even in forensic evaluations that assess psychopathic traits, it is not apparent how

individuals high on psychopathy may respond differently from those low on psychopathy on

other parts of the evaluation. By examining how individuals high on psychopathic personality

traits respond during psychological evaluations, researchers can improve methods for accurately

2

assessing these individuals for things relevant in a forensic evaluation, such as psychopathology

and personality traits.

Psychopathy and Deceitfulness

Deceitfulness is a key behavioral component of psychopathy. Psychopaths frequently use

deception as a means to manipulate others for their own personal gain (Hare, 1993). Some

psychopaths even lie for the thrill of it (Hare, 1993), or for what Eckman (1985) refers to as

“duping delight.” They often show no shame when caught in a lie, but are proud of their ability

to deceive others without being discovered (Hare, 1993).

Psychopaths may also be motivated to use deception in a variety of situations where they

must undergo some sort of forensic psychological evaluation (Rogers & Cruise, 2000). For

example, in both forensic and clinical contexts, individuals may attempt to affect the outcome of

assessments by presenting themselves in an overly negative or positive light. Individuals may

accomplish this by engaging in various forms of response distortion (i.e., malingering or

defensiveness) on psychological assessments (Bagby, Marshall, Bury, Bacchiochi, & Miller,

2006). Psychopaths‟ tendencies to lie and manipulate others may have serious implications for

the validity of forensic evaluations if these individuals are able to successfully represent

themselves in an overly positive or negative light. The broad explanation below of the definition

and measurement of response bias is followed by a discussion of the link between psychopathy

and deception in forensic evaluations is explored further.

Malingering

Feigning bad, often referred to as malingering in light of an external incentive, involves

over-reporting psychological problems in order to appear mentally ill (Rogers, 2008). For

instance, defendants standing trial could attempt to present themselves in an overly negative light

3

to gain the sympathy of a jury or to be declared not guilty by reason of insanity (Iverson,

Franzen, & Hammond, 1995). Personal injury litigants, employees claiming workers‟

compensation benefits, and/or social security disability applicants may over-report or exaggerate

symptoms of mental illness in order to gain compensation (Bagby et al., 2006; Iverson et al.,

1995; Wygant et al., 2009). Prisoners may over-report psychological problems in order to affect

the length of their sentence, placement (i.e., lower security level, or a mental health facility), and

responsibilities in prison (i.e., work detail) (Iverson et al., 1995).

Defensiveness

On the other hand, feigning good, commonly referred to as defensiveness, can involve

under-reporting, minimizing, or denying psychopathology, as well as claiming unlikely virtues

(Bagby et al., 2006; Rogers, 2008). Incarcerated individuals might minimize psychological

problems or aggressive proclivities in order to be granted parole. Parents may deny

psychological problems and endorse unlikely virtues (e.g., minor personal faults and

shortcomings that most people will admit) in order to sway the decision of a judge during child

custody litigation. Job applicants may underreport problems or highlight their strengths in order

to increase the probability of being hired (Sellbom & Bagby, 2008b).

Detection of Response Bias Via Self-Report Inventories

Some psychological test instruments have been developed specifically to assess different

response styles, such as malingering and defensiveness. One example is the Structured Interview

of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992), which uses eight subscales to

measure various types of replies related to an over-reporting response style. The SIRS subscales

focus on a variety of detection strategies, such as: Rare Symptoms (bona fide symptoms that are

infrequently endorsed in clinical populations); Blatant Symptoms (those that are obvious signs of

4

a mental disorder); and Symptom Severity (symptoms which are unbearably severe) (Rogers,

Gillis, Dickens, & Bagby, 1991).

In addition to measures designed solely to assess response styles, a number of widely

used omnibus personality inventories, such as the Millon Clinical Multiaxial Inventory (MCMI;

Millon, 1983, 1987, 1997) and Personality Assessment Inventory (PAI; Morey, 1991, 2007),

employ validity scales to detect various forms of response bias, such as over- and under-

reporting (see Sellbom & Bagby, 2008a, for a review). Of these, the Minnesota Multiphasic

Personality Inventory (MMPI; Hathaway & McKinley, 1940) was one of the first self-report

measures to include validity scales (Baer & Miller, 2002). Its successor, the MMPI-2 (Butcher,

Graham, Ben-Porath, Tellegen, Dahlstrom, & Kaemmer, 2001), includes validity scales designed

to assess over-reporting (e.g., the Infrequency [F], Back F [FB], and Infrequency

Psychopathology [Fp] scales) and under-reporting (e.g., the Lie [L] and Correction [K] scales). A

meta-analysis of the utility of MMPI-2 validity scales in detecting over-reporting indicated mean

effect sizes (Cohen‟s d) of 2.21 (F), 1.62 (FB) and 1.90 (Fp) (Rogers et al., 2003). A meta-

analysis conducted by Baer and Miller (2002) reported a mean effect size of 1.25 (Cohen‟s d) for

MMPI-2 validity scales in detecting under-reporting. The MMPI-2 is one of the most commonly

employed multi-scale inventories utilized in both clinical and forensic settings to assess

personality and psychopathology (Archer, Buffington-Vollum, Stredny, & Handel, 2006;

Camara, Nathan, & Puente, 2000).

The MMPI-2 Restructured Form (MMPI-2-RF; Ben-Porath & Tellegen, 2008) was

designed to capture the same important clinical information as the MMPI-2, but in a more

efficient manner. The MMPI-2-RF is comprised of six sets of scales, which can be derived from

the original MMPI-2 item pool: Higher-Order (H-O), Restructured Clinical (RC), Specific

5

Problem (SP), Interest, Personality Psychopathology Five (PSY-5), and Validity scales. The

eight Validity scales on the MMPI-2-RF include revised versions of MMPI-2 scales (with one

new scale, Fs), which are specifically designed to detect over- and under-reporting of

psychological problems (Gervais, Ben-Porath, Wygant, & Sellbom, 2010; Sellbom & Bagby,

2008b; Sellbom & Bagby, 2010; Sellbom, Toomey, Wygant, Kucharski, & Duncan, 2010;

Wygant et al., 2009; Wygant et al., 2010).

The MMPI-2-RF offers two such validity scales that were designed to detect over-

reporting of psychopathology and will be used in the current study. The Infrequent Responses

(F-r) scale is used as a general indicator of over-reporting of psychological problems (Wygant et

al., 2009). F-r is a 32-item scale containing items endorsed by less than 10% of both men and

women in the normative sample (Ben-Porath & Tellegen, 2008). These include a variety of

psychological, somatic, and cognitive complaints. The 21-item Infrequent Psychopathology

Responses (FP-r) scale is based on the MMPI-2‟s FP scale, which was designed by Arbisi and

Ben-Porath (1995) to help identify over-reporting of severe psychopathology (Sellbom et al.,

2010). FP-r retains 17 items from the original FP scale, along with four additional items which

showed incremental predictive power for over-reporting (Ben-Porath & Tellegen, 2008).

Two MMPI-2-RF scales designed to detect under-reporting on a test instrument will be

used in the current study. The Uncommon Virtues (L-r) and Adjustment Validity (K-r) scales

were developed by Ben-Porath and Tellegen (2008) via factor analyses of the items from the L,

K, S, & Wiggins Social Desirability scales of the original MMPI-2. Items that consistently

loaded on two primary factors, and those that did not cross-load on the opposite factor were

selected for the L-r and K-r scales, which are each comprised of 14 items. Elevated scores on L-r

suggest denial of minor personal faults and shortcomings to which most people will admit.

6

Elevated scores on K-r indicate that the individual presented him- or herself as well-adjusted.

When used to differentiate between under-reporters and honest responders, L-r and K-r have

shown large effect sizes similar to those obtained for the MMPI-2, indicating that L-r and K-r

work just as well as their original counterparts (Sellbom & Bagby, 2008b).

Psychopathy and the Ability to Feign

Now that psychopathy and response bias have been defined and discussed in broad terms,

the link between the two will be further elaborated. Kucharski, Duncan, Egan, and Falkenbach

(2006) reported that inmates with psychopathy scored higher on certain over-reporting indices

(specifically the MMPI-2‟s F and F-K, the Negative Impression Management [NIM] scale of the

PAI, and the SIRS) than did those low on psychopathy, but the former group did not score

significantly higher on other indices (i.e., the MMPI-2‟s FB and Fp, and the PAI‟s Rogers‟

Discriminant Function Scale [RDF; Rogers, Sewell, Morey, & Ustad, 1996] and Malingering

Index [MAL; Morey, 1996]). Of those high on psychopathy, however, a high percentage were

not identified as malingerers. Kucharski et al. (2006) concluded that this provides evidence that

psychopathy should not be a diagnostic indicator of malingering, because not all individuals high

on psychopathy appear to malinger. However, it is also possible that a higher percentage of

individuals high on psychopathic personality traits actually did malinger, but were not detected

and identified as malingerers, as individuals with such traits might be more adept at deception.

Edens, Buffington, Tomicic, and Riley (2001) administered the Psychopathic Personality

Inventory (PPI; Lilienfeld & Andrews, 1996), a self-report measure designed to assess

psychopathic personality traits in non-incarcerated populations, to a sample of college students

(n = 186). These participants first completed the PPI under standard instructions, then under

instructions to present themselves in a positive light. Results indicated that those high on

7

psychopathic personality traits (i.e., those who scored above the mean on the PPI under standard

instructions) significantly lowered their PPI scores when asked to present themselves in a

positive light. However, those low on psychopathic personality traits (i.e., those who scored

below the mean on the PPI under standard instructions) were not able to significantly lower their

PPI when asked to present themselves in a positive light.

Book, Holden, Starzyk, Wasylkiw, and Edwards (2006) administered Levenson‟s Self-

Report Psychopathy Scale (LSRP; Levenson, Kiehl, & Fitzpatrick, 1995), a 26-item self-report

measure of psychopathy, to a sample of college students (n=201) under standard instruction.

Participants were then instructed to either “fake bad” or “fake good” on the Holden

Psychological Screening Inventory (HPSI; Holden, 1996). Individuals in the “fake good” group

who successfully faked good (i.e., scored below cut off on the HPSI‟s validity index) scored

significantly higher on psychopathy than those who were identified as faking good (i.e., scored

above cut off on the HPSI‟s validity index), but those who successfully faked bad did not score

significantly higher on psychopathy than those who were identified as faking bad. These findings

suggest that while individuals high on psychopathic personality traits are better able to present

themselves in a positive light than those low on such traits, there is no difference in level of

psychopathic personality traits for individuals able to successfully present themselves in a

negative light.

In another undergraduate sample (n = 200), MacNeil and Holden (2006) found no

significant differences in level of general psychopathy (as measured by PPI Total score) between

individuals who successfully and unsuccessfully faked good or bad on the HPSI (as classified

using cutoff scores on the HPSI‟s validity index as well as the Impression Management [IM]

scale of the Balanced Inventory of Desirable Responding [BIDR; Paulhus, 1998]). However,

8

participants who avoided detection while faking good scored higher on the Machiavellian

Egocentricity and Blame Externalization and lower on Stress Immunity subscales of the PPI.

This finding indicates that various traits or aspects of the psychopathy construct may have

differential influence on an individual‟s ability to avoid detection when feigning symptoms of

psychopathology.

In a small sample of prison inmates (n = 55), Poythress, Edens, and Watkins (2001)

examined relations between psychopathy (measured by PPI scores) and scores on malingering

indices (i.e., SIRS; PAI‟s NIM, MAL, and RDF; and the Structured Inventory of Malingered

Symptomatology [SIMS; Smith & Burger, 1997]). Using a “known-groups” design, participants

were classified as either malingerers or nonmalingerers by prison staff. Nonmalingerers were

instructed to feign psychosis while completing the malingering indices. Scores from the

malingering and nonmalingering groups were compared to inmates from the general population

(honest responders who had not been classified as malingerers during their admission procedure)

who completed the same measures under standard instructions. Results indicated that known

malingerers who scored high on the PPI were better able than nonmalingerers and honest

responders to avoid detection on the PAI‟s RDF scale; however, known malingerers who were

high on psychopathic personality traits were unable to avoid detection on the PAI‟s NIM scale.

Statement of the Problem

Previously cited investigations have yielded inconsistent results regarding the

associations between psychopathy and response bias (i.e., malingering and defensiveness), and

most of these studies have significant methodological limitations. Some of the previously cited

studies (e.g., Poythress et al., 2001) had small sample sizes, which could have resulted in Type II

errors. Other studies (e.g., Book et al., 2006; MacNeil & Holden, 2006) used measures without

9

well-established utility in detecting response styles (e.g., HPSI Validity Index), which could have

resulted in substantial misclassification of participants. Poythress and colleagues (2001) relied on

previously established clinician classifications of malingering of unknown reliability, instead of

using a standardized methodology, which also could have resulted in misclassification.

Furthermore, all previous studies have used only one measure to index psychopathy, which does

not account for the variety in conceptualization of this construct. Some of these studies used a

within-subjects design (e.g., Edens et al., 2001; MacNeil & Holden, 2006), which provides the

potential for practice effects. Most of these studies (except Poythress et al, 2001) relied on an

analogue simulation design, which Rogers (2008) indicates may decrease generalizability,

because the individuals may not have sufficient motivation to over- or under-report.

The Current Investigation

The goal of this investigation was to examine the association between psychopathy and

response bias. Previous studies have shown inconclusive results, with some indicating that

psychopathy may be linked to increased ability to successfully present themselves in a positive

light (e.g., Book et al., 2006; Edens et al., 2001), but others finding the opposite (e.g., MacNeil

and Holden, 2006). Contradictory evidence has also been found in regard to feigning bad, with

some research indicating that psychopathy is not associated with the ability to successfully fake

bad (e.g., Book et al., 2006;MacNeil & Holden, 2006). Other research indicates that those high

on psychopathy scored higher than those low on psychopathy on measures of malingering (e.g.,

Kucharski et al., 2006;Poythress et al., 2001).

To address the inconclusive results in the literature to date, the current study employed an

innovative design. No study to date has specifically investigated whether psychopathy moderates

the utility of validity scales (i.e., MMPI-2-RF F-r, FP-r, L-r, and K-r scales) in differentiating

10

between individuals who over- or under-report and honest respondents, such that individuals

higher on psychopathic personality traits would be better at avoiding detection than individuals

low on such traits. Such moderation would indicate that those high on psychopathy are more

adept at deception.

The first hypothesis stated that for individuals asked to over-report symptoms of

psychopathology, psychopathy would not moderate the validity scale (F-r and Fp-r) scores‟

utility in differentiating over-reporters from honest responders. Although psychopaths have been

shown to be more prone to engage in deceptive tactics during psychological assessments (Edens

et al., 2001) and intuition might suggest that they would be more successful at doing so because

of practice in deception (MacNeil & Holden, 2006), most of the available literature investigating

this topic (albeit inconsistent) suggests that psychopaths are not better able to feign

psychopathology than non-psychopaths (e.g., Poythress et al., 2001).

The second hypothesis argued that for individuals asked to under-report symptoms of

psychopathology, psychopathy would moderate the validity scale (L-r and K-r) scores‟ utility in

differentiating under-reporters from honest responders more so than for those asked to over-

report symptoms. Some investigations indicate that psychopaths are better able to under-report

than over-report psychopathology (Book et al., 2006; Edens et al., 2001; MacNeil & Holden,

2006). In accordance with such evidence, it was hypothesized that individuals high on

psychopathic personality traits would be better able to under-report symptoms of

psychopathology without being detected than they would be able to over-report

psychopathology.

These hypotheses were tested using two studies, which utilized using two samples drawn

from two different populations with two different methodologies: an analogue simulation design

11

(undergraduate students) and a known-groups design (criminal defendants). Research conducted

using known groups of over-reporters (or those responding defensively) is likely more

generalizable to clinical populations, whereas an analogue simulation design allows for greater

internal control (Rogers, 2008). Randomly assigning individuals to groups and instructing them

to respond in a certain way (e.g., feign good, feign bad, or respond honestly) reduces the

possibility of group differences on possible confounding variables (Rogers, 2008). For example,

if individuals in the over-reporting group were significantly higher on psychopathy than

individuals in the honest group, they may be more motivated to feign in the first place.

The current investigation sought to address some of the limitations from previous

research. These studies included two samples sufficiently large enough for the planned statistical

analyses. Over-reporting (in both studies) and under-reporting (Study 1 only) of

psychopathology was assessed using the MMPI-2-RF Validity scales, which have proven utility

in detecting such response styles (e.g., Gervais et al., 2010; Marion, Sellbom, & Bagby, 2011;

Sellbom & Bagby, 2008b; Sellbom & Bagby, 2010; Sellbom et al., 2010; Wygant et al., 2009;

Wygant et al., 2010). Study 1 used a control group of individuals who completed the MMPI-2-

RF under standard instructions. Both studies utilized a cross-sectional, between-subjects

approach in order reduce practice effects that might have been encountered in previous studies

where participants completed the same inventory under different sets of instructions (e.g., Edens

et al., 2001; MacNeil & Holden, 2006). The first study included multiple means of assessing

psychopathy, in order to obtain an accurate reflection of psychopathic personality traits for each

participant. In contrast, the second study used the well-established Psychopathy Checklist-

Revised (PCL-R; Hare, 1991; 2003). Due to the fact that there is no “gold standard” self-report

psychopathy measure for use in non-incarcerated populations, Study 1 utilized a combined score

12

from several psychopathy measures, to ensure that a comprehensive assessment of psychopathy

was obtained. The PCL-R is the “gold standard” for use with incarcerated samples, and was,

therefore, most appropriate for use in Study 2.

Only one study (MacNeil & Holden, 2006) to date has specifically examined different

facets of psychopathy (as measured by the eight subscales of the PPI) and their relationship to

detection avoidance. The current investigation assessed psychopathy using multiple measures

and examined the three major domains (affective, interpersonal, and behavioral) of psychopathy

to determine if any increased detection avoidance could be attributable to one particular facet.

13

Study 1

Method

This study implemented an analog simulation design to determine if psychopathy

moderates the utility of the MMPI-2-RF‟s F-r, FP-r, L-r, and K-r Validity scales to differentiate

between individuals asked to either over- or under-report from those who took the test honestly.

Participants. Potential participants were 777 undergraduate students. MMPI-2-RF

protocols are generally deemed invalid for interpretation if an individual has produced too many

unscorable responses (i.e., left items blank or marked items both true and false), or if the

individual has engaged in random or indiscriminant fixed responding. Thus, I excluded protocols

with these response styles, based on elevated Cannot Say (CNS) > 17, VRIN-r (>79T) or TRIN-r

(>79T) scores, respectively, as recommended by Ben-Porath and Tellegen (2008). After

excluding MMPI-2-RF profiles based non-content invalid responding, the sample size was

reduced to 627 participants. To determine if those who were excluded differed significantly from

those who were not excluded, demographic information (including baseline psychopathy scores

on each measure) was compared using t-tests (continuous variables) and chi-square tests

(categorical variables). The 149 individuals excluded based on MMPI-2-RF guidelines differed

from those not excluded in regard to ethnicity, χ2 (6, N = 777) = 26.94, p < .001, with those

excluded being more likely to be African American, Hispanic, Asian American Indian, or Alaska

native than those not excluded. However, the effect size for this difference (r = .19) was small

(Cohen, 1988). No other significant differences (p < .01) in demographic information (including

14

baseline psychopathy scores on all measures) between those excluded based on MMPI-2-RF

guidelines and the remaining participants.

Profiles were also excluded if individuals reported inadequate understanding of the study

or poor compliance with instructions, as determined by responses on the post-test questionnaire.

Profiles were excluded if they contained any of the following responses: incorrect answer for

Question 1; “A” for Question 6; or “A”, “B”, “C”, “D”, or “E” for Question 7. Beginning with

the 627 participants remaining after MMPI-2-RF invalid profiles were dropped, excluding based

on Question 1 reduced the sample size to 499, and Question 6 reduced the sample size to 488.

Using all three questions resulted in a final sample size of n = 465. Participants who were

excluded based on each question of the post-test questionnaire did not evidence significant

differences (p < .01) in demographic information (including baseline psychopathy scores on all

measures) as compared to those in the final sample.

After excluding invalid protocols based on MMPI-2-RF validity scales and the MMPI-2-

RF post-test questionnaire, the final sample consisted of male (n = 154) and female (n = 311)

undergraduate students age 18 or older (M = 19.08, SD = 1.84). In regards to ethnicity,

participants were primarily Caucasian (84.3%) or African American (11.0%), with the remaining

4.7% of other or mixed ethnicities. Most participants were freshmen (66.0%) or sophomores

(21.9%). The majority of this sample (98.3%) reported being single and never married. Based on

a power analysis, this sample size of n = 465 is sufficient to give enough statistical power (1 -

= 0.86) to yield a meaningful effect size increment of .02 (∆R2) (Cohen, 1988)

1. Subjects were

randomly assigned to the Feign Good (n = 100), Feign Bad (n = 119) and Honest (n = 246)

conditions.

1 This meaningful effect size involves a Nagelkerke estimated r

2 change based on a logit

function, as I conducted logistic regression analyses.

15

Measures.

Psychopathic Personality Inventory – Revised (PPI-R; Lilienfeld & Widows, 2005). The

original PPI (Lilienfeld & Andrews, 1996) is a 187-item self-report questionnaire designed to

measure psychopathic personality traits in non-incarcerated samples using eight subscales. These

subscales load primarily onto two factors, with the exception of the Coldheartedness subscale

(Benning, Patrick, Hicks, Blonigen, & Krueger, 2003). The Fearless Dominance factor includes

subscales measuring Social Potency, Fearlessness, and Stress Immunity, while the Impulsive

Antisociality factor includes subscales measuring Carefree Nonplanfulness, Impulsive

Nonconformity, Machiavellian Egocentricity, and Blame Externalization. Participants indicate

the degree to which they agree with statements on the following scale:1 (false), 2 (mostly false, 3

(mostly true), or 4 (true). The PPI has well-established reliability and validity (e.g., Lilienfeld &

Hess, 2001; Mullins-Nelson, Salekin, & Leistico, 2006; Vaughn, Newhill, DeLisi, Beaver, &

Howard, 2008). PPI Total Scores have shown moderately high correlations with Total (.54, p <

.001), Factor I (.54, p < .001), and Factor II (.40, p < .01) of Hare‟s (1991, 2003) Psychopathy

Checklist – Revised (PCL-R), the gold standard of psychopathy assessments (Poythress, Edens,

& Lilienfeld, 1998). Various studies have reported internal consistencies (Cronbach‟s alphas)

ranging from .85 to .94 for PPI total scores (Cale & Lilienfeld, 2006; Edens, Poythress, &

Lilienfeld, 1999; Lilienfeld & Andrews, 1996; Lilienfeld & Hess, 2001). The PPI-R was

designed to capture the same information (including all eight subscales) as the PPI, in a shorter,

154-item version. Internal consistency (Cronbach‟s alpha) for the Total score in the current

sample was .93, and internal consistencies for the subscales ranged from .81 (Coldheartedness)

to .89 (Fearlessness). The PPI-R is a copyrighted, proprietary instrument and could therefore not

be included in the appendices.

16

Triarchic Inventory. Patrick‟s (2008) Triarchic Inventory is a 58-item measure designed

to measure psychopathy according to a triarchic conceptualization of psychopathy (Patrick,

Fowles, & Krueger, 2009). This conceptualization seeks to integrate differing historical

perspectives of psychopathy by focusing on recurring themes. The scale emphasizes three

separate facets of psychopathy: boldness, meanness, and disinhibition. The scale is currently in

its infancy, but initial research has shown promising reliability and validity for this measure

(Patrick, 2010). Internal consistencies (Cronbach‟s alpha) for the current sample were: .89

(Total), .84 (Boldness), .88 (Meanness), and .84 (Disinhibition). A copy of the Triarchic

Inventory items are included in Appendix A.

Self-Report Psychopathy Scale – 2nd

Edition (SRP-II). Hare‟s (1985) SRP-II is a self-

report inventory designed to assess the same characteristics of psychopathy as the PCL-R.

Participants respond to each of the 60 items on a scale from 1 (strongly disagree) to 7 (strongly

agree). In addition to a total score, the SRP-II yields four factor scores: Interpersonal, Affective,

Lifestyle, and Antisocial (Williams, Paulhus, & Hare, 2007). The SRP-II has shown moderate to

high correlations with other measures of psychopathy, such as the PCL-R (.54; Hare, 1991) and

the PPI (.77; Williams & Paulhus, 2004). Internal consistency (Cronbach‟s alpha) for the Total

score in the current sample was .88. A copy of the SRP-II items are included in Appendix B.

Levenson’s Self-Report Psychopathy Scale (LSRP). The LSRP (Levenson, Kiehl, &

Fitzpatrick, 1995) is a 26-item self-report inventory of psychopathic personality traits designed

for use in noninstitutionalized populations. In addition to a Total Score, investigations (e.g.,

Brinkley, Diamond, Magaletta, & Heigel, 2008; Sellbom, 2010) have indicated that the LSRP

yields three factor scores (Egocentricity, Callous, and Antisocial) which were examined in the

current study. Brinkley, Schmitt, Smith, and Newman (2001) showed that the LSRP correlates

17

moderately with the PCL-R, and other studies have shown more general support for its construct

validity (Levenson et al., 1995; Lynam, Whiteside, & Jones, 1999; Sellbom, 2010). Internal

consistencies (Cronbach‟s alpha) for the current sample were: .86 (Total), .84 (Factor 1), .68

(Factor 2), and .65 (Factor 3). A copy of the LSRP items are included in Appendix C.

Minnesota Multiphasic Personality Inventory – Restructured Form (MMPI-2-RF). The

MMPI-2-RF (Ben-Porath & Tellegen, 2008) consists of 338 items derived from the 567-item

MMPI-2 (Butcher, et al., 2001). Participants select either “true” or “false” for each item, as it

applies to them. Ben-Porath and Tellegen (2008) provide extensive reliability and validity

information for the MMPI-2-RF in the technical manual. For this study, the F-r, Fp-r, L-r, and K-

r scales were employed. All scales have shown significant correlation with their MMPI-2

counterparts in a variety of samples, with correlation coefficients above .90 for F-r, Fp-r, and L-r,

and ranging from .84-.88 for K-r (Sellbom & Bagby, 2008b; Sellbom et al., 2010). F-r and FP-r

have significantly differentiated between malingerers and non-malingerers in multiple samples,

with large effect sizes (Cohen‟s d) ranging from 2.11-2.37 for F-r and 2.07-2.34 for FP-r

(Sellbom et al., 2010). L-r and K-r have significantly differentiated under-reporters from honest

responders across multiple samples, with large effect sizes (Cohen‟s d) ranging from .65-1.13

and .76-1.44, respectively (Sellbom & Bagby, 2008). The MMPI-2-RF is a copyrighted,

proprietary instrument and could not be included in the appendices.

MMPI-2-RF Post-Test Questionnaire. This 10-item questionnaire was designed for this

study. It was designed to assess participants‟ understanding of the study and their level of

compliance with instructions (see Appendix D).

Procedure. I recruited participants at The University of Alabama through the Department

of Psychology‟s subject pool. Students in introductory psychology courses are required to either

18

participate through subject pool research or complete an alternative assignment for course credit.

Participants signed up for the study online and reported to Room 408, Gordon Palmer Hall, at the

specified date and time. Participation took approximately two hours, and students received three

hours of research credit, in accordance with departmental policy. It was expected that, as

students admitted to the University of Alabama, participants would have the ability to read at or

above a 5th

grade reading level (Flesch-Kincaid Grade Level; Flesch, 1949). This is the highest

reading level required to complete any of the questionnaires (i.e., the MMPI-2-RF; Pearson

Education, Inc., 2009).

Before participants arrived, the researcher set up questionnaire administration on lab

computers and entered an ID number into each computer. Numbers were given in sequential

order, beginning with 1000 (“Honest” condition), 2000 (“Under-reporting” condition), or 3000

(“Over-reporting” condition). All participants in each session were in the same condition, and

order of condition administration was randomized. No personal or identifying information was

collected from participants and ID numbers were not linked to the participants in any way. ID

numbers were used only as a reference to link together the data from the questionnaires

completed by each participant.

Participants were informed of their rights as research participants, the purpose of the

study, and what would be expected of them should they choose to participate. They were

informed that they could complete an alternative assignment to receive credit for class if they

declined to participate in this study, and that they were free to leave the study at any time without

penalty. Participants were told that they may be asked to respond in a manner that does not

accurately reflect their personality, but that they would receive specific instructions if asked to

do so. They were advised that their results will remain anonymous and kept protected by the

19

primary investigator. Participants experienced no risks greater than those experienced during

everyday living.

After providing informed consent, participants were administered a series of self-report

questionnaires. All participants completed the PPI, SRP-II, LSRP, and Triarchic Inventory under

standard instructions. They were then presented with instructions for the MMPI-2-RF. The

“Honest” condition was presented with standard instructions; the “Under-reporting” condition

was presented with instructions to present themselves in a positive light; and the “Over-

reporting” condition was presented with instructions to present themselves in a negative light.

Instructions for the “Under-reporting” and “Over-reporting” conditions are included in

Appendices E and F, respectively. Upon finishing the MMPI-2-RF, participants completed a 10-

item questionnaire designed specifically for this study to assess their understanding of the study

and their level of compliance with instructions (see Appendix D). Once participants completed

the self-report measures, they were debriefed. The researcher addressed any questions or

concerns they had and provided contact information for the Primary Investigator, Faculty

Advisor, and the University of Alabama‟s Office of Research Compliance.

Results

Calculation of Psychopathy Scores. To obtain psychopathy factor scores for each

participant, I first conducted an exploratory factor analysis (EFA) using maximum likelihood

estimation to extract the optimal number of factors. For this analysis, I entered all subscales (the

eight subscales of the PPI, the three factor scales of the Triarchic Inventory, and the three factor

scales of the LSRP) from the psychopathy measures used in this study, examining factor

solutions in which all factors met the Kaiser (1960) criterion of an eigenvalue of 1.00 or greater.

I considered a maximum of four factors, as no previous empirical research has found support for

20

a factor model of psychopathy that extends beyond four factors. Eigenvalues for the first four

unrotated factors were 5.04, 2.59, 1.48, and .97, respectively. A two-factor structure appeared to

assess affective-impulsive features (Factor 1) and bold/fearless features (Factor 2). In regard to

Factor 1, it would make little conceptual sense for scales assessing a cold, mean affective style

and scales measuring a carefree or disinhibited personality to be included on the same factor, as

these sets of traits typically load on different factors on virtually all psychopathy measures (e.g.,

PCL [Hare, 1991; 2003], LSRP [Levenson, Kiehl, & Fitzpatrick, 1995; Sellbom, 2010], SRP-II

[Hare, 1985; Williams, Paulhus, & Hare, 2007], and Antisocial Process Screening Device

[APSD][Frick & Hare, 2001; Sadeh, Verona, Javdani, & Olson, 2009]). With this two-factor

structure, all three factors of the LSRP were included on Factor 1, and only four subscales loaded

on Factor 2. Thus, this two-factor model did not appear to differentiate well between the various

components of psychopathy. A four-factor structure was not considered, as only three factors

met the Kaiser (1960) criterion of an eigenvalue of 1.00. The three-factor model was

conceptually most defensible and was therefore chosen. Subscales loading on the first rotated

factor captured content indicative of disinhibitory and impulsive personality characteristics;

those loading on the second factor measured a mean or coldhearted interpersonal style; and those

loading on the third factor assessed content indicative of a bold or fearless personality. This

factor structure lined up well with the Triarchic model of psychopathy (e.g., Patrick, 2010),

which includes the three domains of Disinhibition, Meanness, and Boldness. Indeed, each of the

three factors were anchored by one of these scales, and the additional subscales loading on each

factor capture similar or relevant material to each of these domains. The resulting structure was

rotated according to an oblique promax method. Table 1 displays the results of the EFA,

indicating which subscales loaded on each of the three factors.

21

Table 1.

Eigenvalues and Factor Loadings for Exploratory Factor Analysis of Psychopathy Scales

Disinhibition

(35.96%)

Meanness

(18.49%)

Boldness

(10.57%)

Triarchic Disinhibition .83

LSRP Factor 3 .77

PPI-R Blame Externalization .61

PPI-R Rebellious Nonconformity .57

PPI-R Carefree Nonplanfulness .53

PPI-R Coldheartedness -.32 .95

Triarchic Meanness .78

LSRP Factor 1 .63

LSRP Factor 2 .61

PPI-R Machiavellian Egocentricity .40 .47

Triarchic Boldness 1.01

PPI-R Social .75

PPI-R Stress Immunity -.30 .62

PPI-R Fearless Dominance .54

Note. Only factor loadings equal to or greater than .30 are shown.

I used a cutoff of .30 for factor loadings to determine if a scale loaded on a factor.

Subscales that loaded on more than one factor were only included in the factor score for which

they had the highest loading. To generate scores for each participant on overall psychopathy as

well as the factor scores, I standardized and averaged the scores on each subscale (i.e., an

average of z-scores for subscales comprising the respective factors) for each participant.

22

Group Differences in Validity Scale and Psychopathy Scores. I next conducted one-

way Analyses of Variance (ANOVA) using three groups (“Over-reporting”, “Under-reporting”,

and “Honest”) to determine if group differences were present for validity scale (F-r, Fp-r, L-r,

and K-r) and psychopathy scores (Total score, Meanness factor, Disinhibition factor, and

Boldness factor). Cohen‟s d effect size estimates were used to characterize meaningful

differences, with .20 indicating a small effect, .50 a medium effect, and .80 a large effect (Cohen,

1988). Table 2 displays these results.

Table 2.

Means, Standard Deviations, F-tests, and Effect Sizes for Honest, Under-reporting, and Over-

reporting Groups

Honest Under-

reporting

Over-

reporting

F d1 d2 d3

(n = 246) (n = 100) (n = 119)

F-r 58.73a

(14.09)

49.63b

(8.64)

148.36c

(30.00)

786.71*** .66 -3.70 -3.48

FP-r 61.15a

(8.18)

56.88a

(5.91)

159.35b

(23.64)

669.93*** .35 -3.46 -3.05

L-r 52.40a

(9.66)

59.65b

(14.01)

52.40a

(11.11)

17.48*** -.67 .02 .60

K-r 46.04a

(8.93)

52.56b

(9.97)

34.72b

(6.53)

125.90*** -.71 1.39 2.18

Disinhibition .15

(3.51)

.27

(4.16)

-.53

(3.61)

1.71 -.03 .19 .21

Meanness .19

(3.92)

-.20

(4.65)

-.23

(3.88)

.58 .09 .11 .01

Boldness .18

(3.09)

-.30

(3.67)

-.13

(3.11)

.97 .15 .10 -.05

Psychopathy

Total

.21

(3.38)

-.13

(4.23)

-.33

(3.51)

1.00 .09 .16 .05

Note. Means with different subscripts are significantly different at p < .05. d1 = Effect size for

Honest vs. Under-reporting groups, d2 = Effect size for Honest vs. Over-reporting groups, d3 =

Effect size for Over-reporting vs. Under-reporting groups. *** p <.001.

23

As expected, scores on F-r and Fp-r were significantly higher in the “Over-reporting”

group relative to the other groups (all Cohen‟s ds > 3.04), whereas scores on L-r and K-r were

significantly higher in the “Under-reporting” group relative to the other groups (all Cohen‟s ds >

.59). The over-reporting scales (F-r and Fp-r) differentiated over-reporters from honest

responders very well (Cohen‟s d = -3.70 and -3.46, respectively). The under-reporting scales (L-r

and K-r) also differentiated under-reporters from honest responders well (Cohen‟s d = -.67 and -

.71, respectively), though the effect size estimates were not nearly as high as those for the over-

reporting scales. Also as expected, the “Over-reporting”, “Under-reporting”, and “Honest”

groups were not significantly different on any of the psychopathy scores, due to random

assignment of participants to groups.

Psychopathy as a Moderator. I next conducted a series of hierarchical binomial logistic

regression analyses to determine if psychopathy moderated the ability to avoid detection by the

MMPI-2-RF validity scales. Table 3 displays the results of these analyses.

24

Table 3.

Hierarchical Logistic Regression Analyses Predicting Over-reporting vs. Not Over-reporting, or

Under-reporting vs. Not Under-reporting, Using Validity Scale, Psychopathy, and Psychopathy

X Validity Scale.

Full Model

Psychopathy Score F-r Psychopathy F-r X Psychopathy R2 Δ R

2

2 chg

Disinhibition .42*** -.24** -.01 .85 .00 .89

Meanness .39*** -.16* .01 .84 .00 .24

Boldness .38*** -.13 .02 .84 .00 2.08

Total .40*** -.23** .01 .85 .00 .39

Full Model

Fp-r Psychopathy Fp-r X Psychopathy R2 Δ R

2

2 chg

Disinhibition .63*** -.12 .02 .83 .00 .62

Meanness .65*** -.10 .04 .83 .00 3.27

Boldness .64*** -.11 .04 .83 .01 2.95

Total .68*** -.15* .05* .83 .01 4.07*

Full Model

L-r Psychopathy L-r X Psychopathy R2 Δ R

2

2 chg

Disinhibition .29*** .05 .01 .12 .00 .50

Meanness .28*** -.03 .02 .12 .01 1.98

Boldness .26*** -.03 -.03 .12 .00 1.81

Total .28*** -.01 .02 .12 .01 1.15

Full Model

K-r Psychopathy K-r X Psychopathy R2 Δ R

2

2 chg

Disinhibition .28*** .04 .02 .16 .01 2.91

Meanness .26*** -.10* .04** .17 .04 9.58**

Boldness .30*** -.11* -.02 .18 .01 1.40

Total .26*** -.09 .03* .15 .02 5.47*

Note. Nagelkerke R2 estimation was used for logistic regression. Δ R

2 = change in R

2 between

Step 1 (validity scale and psychopathy scale entered) and Step 2 (validity scale X psychopathy

25

scale interaction entered). 2 chg = change in

2 between Step 1 and Step 2. * p < .05, ** p < .01,

*** p < .001.

Separate regression analyses were conducted for the four MMPI-2-RF scales and the

psychopathy scores (total and three factors). For this purpose, I created centered interaction

terms. For example, for the interaction term between F-r and the total psychopathy score, I

subtracted the mean F-r score from each participant‟s F-r score, and subtracted the mean total

psychopathy score from each participant‟s total psychopathy score. I then multiplied these scores

together to produce a centered interaction term. This method helps reduce the potential for

multicollinearity (correlation between the interaction term and the main effects variables)(e.g.,

Cohen, Cohen, West, & Aiken, 2003).

In terms of over-reporting, one of the two MMPI-2-RF over-reporting scales (F-r and FP-

r) and the psychopathy score (total or factor score) were entered in the first step, and the centered

interaction term (i.e., cross-product) of the psychopathy score and MMPI-2-RF scale score in the

second step, as predictors for the dichotomous outcome of either “Over-reporting” or “Not Over-

reporting.” The same procedure was used for under-reporting analysis, except the MMPI-2-RF

L-r and K-r scales were used. Any a moderation effect was deemed present by a statistically

significant increase in model fit (i.e., 2 change) and a Nagelkerke-estimated ∆R

2 > .02. Because

there is no standard set for a meaningful incremental change in a nonparametric test such as

logistic regression, I used Cohen‟s (1988) recommendation that this level of incremental change

is meaningful in ordinary least squares estimated regression.

I first hypothesized that psychopathy would not moderate the validity scale (F-r and Fp-r)

scores‟ utility in differentiating between over-reporters and honest responders. For the over-

reporting analyses, the interaction terms were not associated with meaningful incremental change

in predictive utility, indicating that psychopathy did not moderate the validity scale scores‟ utility

26

in differentiating over-reporters from honest responders. These findings support my first

hypotheses, insofar that for individuals asked to over-report psychopathology, psychopathy did

not moderate the validity scale (F-r and Fp-r) scores‟ utility in differentiating over-reporters from

honest responders.

My second hypothesis stated that psychopathy would moderate the validity scale (L-r and

K-r) scores‟ utility in differentiating under-reporters from honest responders more so than for

those asked to over-report symptoms. For the under-reporting scales, no meaningful incremental

change was present for L-r. Only two regressions, both for the K-r scale, met the criteria for a

meaningful incremental change provided by the interaction term between validity scale and

psychopathy score. This interaction evidenced incremental utility when Meanness and Total

psychopathy scores (both ΔR2 > .02, both

2 chg significant at p < .05) were predictors,

indicating that psychopathy did moderate the validity scale scores‟ utility in differentiating

under-reporters from honest responders. Figure 1 displays the interaction between Meanness and

K-r when predicting individuals‟ group membership (under-reporting vs. honest).

Figure 1.

Interaction Between Meanness and K-r Score in Predicting Honest vs. Feign Good Groups.

27

These results do not support my second hypothesis. Although moderating effects were

found for this scale, they were actually in the opposite direction from what was expected. That is,

psychopathy (with an emphasis on the Meanness factor) was associated with a poorer ability to

avoid detection by the K-r validity scale. This counterintuitive effect will be considered in detail

later.

28

Study 2

The primary goal of this study was to replicate a portion of the results from Study 1 using

an alternative methodology and setting. This study implemented a known-groups design to

determine if psychopathy moderates the utility of the MMPI-2-RF F-r and FP-r Validity scales to

differentiate between individuals who engaged in over-reporting from those who approached the

test honestly. Because this study utilized a sample of convenience, I was only able to investigate

over-reporting in this study.

Method

Participants and procedure. I used an archival sample of 122 male criminal defendants

who were referred by Federal court between 1994 and 2004 for a forensic psychological

evaluation (i.e., competency to stand trial, criminal responsibility, or aid-in-sentencing). This

sample has been used in previous studies examining the MMPI-2 and MMPI-2-RF Validity

scales (Sellbom et al., 2010; Toomey, Kucharski, & Duncan, 2009). Although I recognize that

this sample size is not powerful enough to detect a small increment in 2 change, previous

research has indicated that MMPI-2-RF validity scales are associated with very large effect sizes

in differentiating between over-reporting and honest individuals in this sample (i.e., Cohen‟s ds >

2.00; Sellbom et al., 2010). Thus, it is reasonable to expect that an orthogonalized cross-product

that involves scales with such large effect sizes would be associated with a greater than

minimum meaningful effect, if present.

Each participant completed the MMPI-2, SIRS, and PCL-R as part of their evaluation. In

a similar fashion as Study 1, I excluded individuals with invalid MMPI-2-RF profiles, based on

29

the same guidelines described for Study 1 (i.e., CNS > 17, VRIN-r > 79T, or TRIN-r > 79T

indicates an invalid profile), which resulted in a final sample size of 99. Participants ranged from

20 to 64 years of age (M = 37). In regards to ethnicity, participants were primarily Caucasian

(58%) or African American (39%), with the remaining 3% of Hispanic or other origin. Ninety-

one percent had previous felony convictions. In regard to mental health treatment, 67% reported

receiving previous mental health treatment, and 44% had at least one previous psychiatric

hospitalization. The most common primary diagnosis was Antisocial Personality Disorder (35%).

Participants were classified as “Over-reporting” or “Not Over-reporting” based on the

SIRS manual guidelines (Rogers et al., 1992). Participants who scored in the “probable” range

on 3 or more subscales on the SIRS, or in the “definite” range on 1 or more subscales, were

classified as “Over-reporting.” The remaining participants were classified as “Not Over-

Reporting.” I examined PCL-R Total scores, as well as scores for Factors I and II and Facets 1-4.

Measures.

Psychopathy Checklist – Revised (PCL-R). Hare‟s (1991, 2003) PCL-R is considered

the “gold standard” of psychopathy assessment (Lilienfeld & Fowler, 2006). It involves a semi-

structured interview format, where the examiner rates the individual on 20 items indicative of

psychopathic characteristics. In addition to a Total Score, the PCL-R yields two factor scores;

Factor I assesses affective-interpersonal characteristics, while Factor II indexes social deviance

aspects of psychopathy (e.g., impulsivity, sensation-seeking, and lack of responsibility). Hare

(2003) has also described four facets underlying Factors I and II. Facets 1 and 2 separate Factor I

into affective and interpersonal features, whereas Facets 3 and 4 separate Factor II into lifestyle

instability and antisocial behavior features, respectively. Internal consistencies (Cronbach‟s

30

alpha) for the current sample were: .91 (Total), .89 (Factor I), .83 (Factor II), .74 (Interpersonal

facet), .90 (Affective facet), .61 (Antisocial facet), and .79 (Lifestyle facet).

SIRS. The SIRS (Rogers et al., 1992) is a structured interview that contains 172 items

designed to index malingering and over-reporting. Four classifications can be made (honest,

indeterminate, probable faking, and definite faking) from each of the following eight scales: Rare

Symptoms, Improbable and Absurd Symptoms, Symptom Combinations, Blatant Symptoms,

Subtle Symptoms, Symptom Severity, Symptom Selectivity, and Reported vs. Observed

Symptoms. The SIRS has shown utility as an external criterion measure in other known-groups

studies examining over-reporting (e.g., Edens, Poythress, & Watkins-Clay, 2007; Sellbom et al.,

2010). Previous studies have indicated that the SIRS is reliable, with internal consistencies

ranging from .77 to .96, and inter-rater reliability coefficients ranging from .97 to 1.00 (e.g.,

Rogers et al., 1992; Vitacco, Rogers, Gabel, & Munizza, 2007).

MMPI-2-RF. For this sample, I scored the MMPI-2-RF validity scales using data from

MMPI-2 administrations. Scores obtained from administration of the 338-item MMPI-2-RF

booklet are interchangeable with MMPI-2-RF scores derived from administration of the 567-

item MMPI-2 booklet (Ben-Porath & Tellegen, 2008; Sellbom et al., 2010).

Results

Group Differences in Validity Scale and Psychopathy Scores. As with Study 1, I

began with one-way ANOVAs to examine group differences in over-reporting validity scale

scores (F-r and Fp-r) and psychopathy scores between “Over-reporting” and “Not Over-

Reporting” groups. Table 4 displays the results of these analyses.

31

Table 4.

Means, Standard Deviations, F-tests, and Effect Sizes for Honest and Over-reporting Groups

Honest Over-reporting F d

(n = 78) (n = 21)

F-r 8.55

(6.51)

21.71

(5.39)

72.31*** -2.09

FP-r 3.14

(2.73)

9.76

(3.85)

80.93*** -2.21

PCL-R Total 22.80

(10.10)

27.90

(7.76)

4.62* -.53

PCL-R Factor 1 8.96

(5.05)

11.38

(3.26)

4.31* -.51

PCL-R Factor 2 11.17

(4.33)

13.03

(3.83)

3.19 -.44

PCL-R Interpersonal

Facet

4.04

(2.83)

5.05

(1.66)

2.44 -.38

PCL-R Affective Facet 5.05

(2.71)

6.38

(1.94)

4.44* -.52

PCL-R Antisocial Facet 5.29

(2.54)

6.43

(2.75)

3.23 -.44

PCL-R Lifestyle Facet 7.68

(2.96)

8.57

(2.23)

1.65 -.32

Note. * p < .05, ** p < .01, *** p <.001.

As expected, scores on F-r and Fp-r were significantly higher in the “Over-reporting”

group relative to the honest group, with large effect size estimates associated with these

differences (both Cohen‟s ds > 2.08). Such differences were expected given that Sellbom et al.

(2010) used part of the current sample to examine the utility of the MMPI-2-RF validity scales in

differentiating between SIRS-defined groups of over-reporting and non-over-reporting

individuals. A few differences were present in psychopathy scores as well, such that individuals

in the “Over-reporting” group had significantly higher PCL-R Total, Factor 1, and Affective

Facet scores compared to those in the “Honest” group (all Fs > 4.30, all ps < .05). Unlike in

32

Study 1, these individuals were not randomly assigned to groups or asked to over-report

symptoms of psychopathology. Given the forensic context in which these individuals were

evaluated, it makes intuitive sense that individuals who malingered are more likely to be

psychopathic.

Psychopathy as a Moderator. Next, I conducted a series of hierarchical binomial

logistic regression analyses to determine if psychopathy acts as a moderator for the ability to

avoid detection by the MMPI-2-RF validity scales. Table 5 displays the results for these

regression analyses.

Table 5.

Hierarchical Logistic Regression Analyses Predicting Over-reporting vs. Not Over-reporting

Using Validity Scale, Psychopathy, and Psychopathy X Validity Scale.

Full Model

PCL-R Score F-r Psychopathy F-r X

Psychopathy

R2 Δ R

2

2 chg

Total .26*** .00 .01 .60 .01 .67

Factor 1 .26*** .06 .01 .62 .01 .83

Factor 2 .26*** .00 .01 .59 .01 .12

Interpersonal Facet .27*** .12 .02 .62 .01 1.06

Affective Facet .26*** .07 .02 .61 .00 .81

Antisocial Facet .26*** -.03 .02 .59 .00 .52

Lifestyle Facet .27*** -.07 .00 .58 .00 .01

Full Model

Fp-r Psychopathy Fp-r X

Psychopathy

R2 Δ R

2

2 chg

Total .47*** .02 .00 .57 .00 .12

Factor 1 .47*** .07 .01 .58 .00 .05

Factor 2 .48*** .00 .02 .57 .00 .42

Interpersonal Facet .48*** .13 .02 .58 .00 .15

Affective Facet .48*** .13 .00 .57 .00 .01

Antisocial Facet .49*** .00 .03 .58 .01 .73

Lifestyle Facet .50*** -.02 .00 .57 .00 .01

Note. Nagelkerke R2 estimation was used for logistic regression. Δ R

2 = change in R

2 between

Step 1 (validity scale and psychopathy scale entered) and Step 2 (validity scale X psychopathy

scale interaction entered). 2 chg = change in

2 between Step 1 and Step 2. * p < .05, ** p < .01,

*** p < .001.

33

Separate regression analyses were conducted for the two over-reporting MMPI-2-RF

scales (F-r and Fp-r) and the PCL-R scores (total, factor, and facet scores). One of the two

MMPI-2-RF over-reporting scales (F-r and FP-r) and the PCL-R score (total, factor, or facet)

were entered in the first step, and the centered interaction term (i.e., cross-product) between

PCL-R score and MMPI-2-RF scale in the second step, as predictors of the dichotomous

outcome of either “Over-reporting” or “Not Over-reporting.” As with Study 1, a moderation

effect was indicated by a statistically significant increase in model fit (i.e., 2 change) and a

Nagelkerke-estimated ∆R2 > .02, as recommended by Cohen (1988) for ordinary least squares

estimated regression. As in Study 1, these results indicate that psychopathy did not moderate the

validity scale scores‟ utility in differentiating over-reporters from honest responders.

34

General Discussion

The goal of this investigation was to examine the relation between psychopathy and

response bias. More specifically, I investigated whether psychopathy acts as a moderator for the

ability to avoid detection by the MMPI-2-RF validity scales when individuals over- or under-

report symptoms of psychopathology, such that individuals high on psychopathic traits would be

better able to avoid detection than individuals low on these traits.

The previous literature on this subject has been inconclusive thus far. Some studies have

indicated that individuals high on psychopathic traits are better than individuals low on these

traits at presenting themselves in a positive light (e.g., Book et al., 2006; Edens et al., 2001),

though one study by MacNeil and Holden (2006) found that individuals who successfully faked

good scored higher on a few specific facets of psychopathy. Others have found that individuals

high on psychopathic traits are not better able than individuals low on these traits to present

themselves in a negative light (e.g., Book et al., 2006, MacNeil & Holden, 2006). However,

other studies have found that individuals high on psychopathic traits score higher than those low

on these traits on some measures of malingering (e.g., Kucharski et al., 2006, and Poythress et

al., 2001). Overall, these results do not show a general consensus regarding whether individuals

high on psychopathic personality traits are better able to successfully feign either good or bad

than those low on these traits.

In light of the majority of evidence from the literature previously cited, I had first

hypothesized that psychopathy would not moderate the validity scale (F-r and Fp-r) scores‟ utility

in differentiating between over-reporters and honest responders. The results from both Study 1

35

and Study 2 support my first hypothesis. When individuals were asked to over-report symptoms

of psychopathology in Study 1, psychopathy did not moderate the utility of the MMPI-2-RF

validity scales in differentiating between over-reporters and honest responders. In Study 2,

psychopathy did not moderate the utility of the MMPI-2-RF validity scales in differentiating

between known malingerers and honest responders (as classified by the SIRS). Individuals

higher on psychopathic personality traits were not better able than individuals low on these traits

to avoid detection on the MMPI-2-RF validity scales.

These findings are consistent with those of Book et al. (2006), who found that individuals

who successfully faked bad did not score significantly higher on psychopathy than those who

were detected. The current findings are also consistent with those from Kucharski et al. (2006),

who found that inmates high on psychopathy did not score significantly higher than those low on

psychopathy on the MMPI-2‟s Fp scale, but did score higher on the MMPI-2‟s F scale than did

those low on psychopathy. Thus, while individuals high on psychopathic traits may attempt to

malinger more frequently than those low on these traits, they are not more successful at doing so.

The current findings partially contradict those from Poythress et al. (2001), who found that

malingerers high on psychopathic traits were able to avoid detection on the PAI‟s RDF scale, but

not on the NIM scale. The contradictory results between these two studies and the current

investigation could be due to methodological differences, such as use of different measures. For

example, the NIM scale is very similar to Fp-r in its development (e.g., Morey, 2007), but the

RDF scale is not similar to any of the MMPI-2-RF validity scales. As described earlier, F-r and

Fp-r were constructed by examining specific items that were infrequently endorsed by normal

and psychiatric populations. The RDF was developed by conducting a discriminant function

analysis to look at 20 scales of the PAI and determine whether or not an individual‟s pattern is

36

similar to that of a bona fide patient (Rogers, Sewell, Morey, & Ustad, 1996). This difference in

detection strategies between the MMPI-2-RF scales and the PAI‟s RDF could explain the

inconsistent findings between Poythress et al. (2001) and the current study. It should be noted,

however, that some research has indicated that the RDF is of questionable validity in forensic

settings (Hawes & Boccaccini, 2009; Sellbom & Bagby, 2008a); thus, the results that use better

established validity scales such as NIM and Fp-r are more likely to be trusted at this point.

My second hypothesis was that psychopathy would moderate the validity scale (L-r and

K-r) scores‟ utility in differentiating under-reporters from honest responders more so than for

those asked to over-report symptoms. In Study 1, I found results which contradict my second

hypothesis. When individuals were asked to under-report symptoms of psychopathology,

psychopathy did not moderate the utility of the MMPI-2-RF‟s L-r validity scale in differentiating

between under-reporters and honest responders. However, psychopathy did moderate the utility

of the MMPI-2-RF‟s K-r scale in differentiating between under-reporters and honest responders.

I had hypothesized that this moderation effect would indicate that individuals high on

psychopathic traits would be better able than those low on these traits to under-report symptoms

of psychopathology without being detected; however, the results indicated that this moderation

effect was in the opposite direction. Individuals low on certain psychopathic personality traits

(e.g., meanness, callousness, coldheartedness) were actually better able than those high on such

traits to avoid detection on the MMPI-2-RF‟s K-r scale. These results contradict Edens et al.

(2001) and Book et al.‟s (2006) findings that those who successfully faked good were

significantly higher on psychopathic traits than were those who were detected. The results of the

current study also contradict MacNeil and Holden‟s (2006) findings that individuals who

successfully faked good were higher on Machiavellian egocentricity and blame externalization,

37

and lower on stress immunity, than individuals who were detected. It is possible that these results

contradict those from the current investigation due to the use of measures without well-

established utility in detecting response bias (e.g., the HPSI used in both Book et al.[2006] and

MacNeil and Holden [2006]). Holden and Evoy (2005) reported that the HPSI validity scale

correlated strongly (-.59 to -.88) with the Social Desirability scale of the Personality Research

Form (PRF; Jackson, 1984), which more closely resembles the L-r scale than the K-r scale. Thus,

these two different methods of assessing feigning good could account for the differences in

findings between MacNeil and Holden (2006) and the current study. In addition, MacNeil and

Holden (2006) classified individuals as “successful” or “not successful” fakers based on cut

scores. It is possible that this method resulted in misclassification of individuals. Also, MacNeil

and Holden (2006) conducted t-tests to determine significant differences in mean psychopathy

scores between faking and honest groups. While they did find some significant results (e.g., for

Machiavellian egocentricity, blame externalization, and stress immunity), the effect sizes were

generally small to moderate (e.g., ds = .40-.60), and not consistent across any one domain of the

psychopathic personality (i.e., in the current study, each of these PPI scales loaded on three

separate domains). Another possible reason for the contradictory results is the use of different

assessments of psychopathy. It is possible that my three-factor model captures a more

comprehensive conceptualization of psychopathy as compared to the measures used in these

other studies, especially with regard to the mean, cold characteristics of psychopathy (i.e., the

PPI does not fully capture meanness; see Patrick, Fowles, & Krueger, 2009).

One possible explanation for the current findings for K-r could be that “mean” or

coldhearted psychopaths might have more difficulty feigning psychological adjustment than do

individuals low on such psychopathic traits. Perhaps the emotional detachment and poor

38

perspective-taking ability that individuals high on this psychopathy factor exhibit make it

difficult for them to take the perspective of what emotional adjustment look like in others. Thus,

when attempting to appear more psychologically adjusted than they really are, they over-

compensate in their presentations and therefore score higher on K-r while doing so. Further

research is needed to elaborate on these hypotheses.

As discussed earlier, psychopathy was not associated with any better or worse ability to

avoid detection by the L-r scale, which assesses unlikely virtues, that is, common and obvious

personal flaws that most individuals would be willing to endorse. These are potentially easier to

feign successfully given the fact that this is a less sophisticated and more superficial method of

feigning. Indeed, Baer and Miller (2002) demonstrated in a meta-analysis on the MMPI-2

validity scales that when individuals are coached on the validity scales, L is associated with a

much smaller effect size in differentiating under-reporting from honest participants relative to K.

Thus, individuals who score high on the meanness domain of psychopathy are more likely to

avoid detection by a scale that focuses on uncommon virtuous than psychological adjustment.

Overall, when integrating the results of the present investigation with the previous

literature, the preponderance of well-designed studies tend to show that individuals high on

psychopathic traits are no better able to avoid detection on validity scales than individuals low on

these traits in terms of over-reporting. In terms of under-reporting, the combined evidence

suggests that individuals high on psychopathic traits are no better able than individuals low in

such traits to avoid detection on validity scales when attempting to make themselves appear

socially desirable. However, specific features of psychopathy may affect an individual‟s ability

to successfully feign psychological adjustment without being detected, such that some features

(e.g., meanness) decrease this ability, while others (e.g., stress immunity) may increase it.

39

The current findings pose implications for forensic psychological evaluations. In general,

individuals high on psychopathic traits are not better able to present themselves in an overly

positive or negative light than those low on these traits, which is an encouraging finding. When

individuals do try to present themselves in an inaccurate light, the MMPI-2-RF validity scales

evidenced utility in distinguishing between those feigning and those who responded honestly.

Because individuals in forensic evaluations often have an incentive to respond in an overly

positive or negative manner (e.g., Rogers & Cruise, 2000), it is important that response bias be

assessed in forensic evaluations to determine whether or not an individual is trying to provide an

inaccurate picture of him- or herself. Given the information that individuals high in psychopathic

traits may be more likely to malinger, it is especially important that these individuals be assessed

for response bias. Individuals high on psychopathy may not be especially adept at over- or

under-reporting psychological symptoms. However, if they engage in response bias on a more

frequent basis than those low on psychopathic traits, the odds are that more individuals high in

psychopathy will avoid detection on validity indices in forensic evaluations. This could result in

a higher number of psychopathic individuals receiving favorable verdicts (e.g., not guilty by

reason of insanity, custody, workers compensation benefits, etc.) and sentences (e.g., mental

health facility vs. general population). Because individuals high on psychopathic traits often

commit more violent crimes than those low on these traits, this could pose a significant threat to

others (e.g., mental health workers and patients). It could also pose a significant financial burden

(e.g., to insurance companies and taxpayers) to provide benefits, treatment, and housing for these

individuals.

Though I have made significant efforts to address some of the methodological issues

found in previous studies, the current investigation is nonetheless also associated with some

40

limitations. In regards to the under-reporting analyses, due to the fact that Study 2 utilized a

sample of convenience, data were not available to re-test my second hypothesis in a second

under-reporting sample. In regards to the over-reporting analyses, the sample in Study 2 was

small and may not have been sufficiently powerful to detect a statistically significant moderation

effect. However, the results were entirely consistent with Study 1 and the extant literature, and

the actual effect size (which is not dependent on sample size) did not reach a meaningful level.

Study 1 utilized a sample of convenience (university students), which limits the generalizability

of these results to other samples. This sample was reduced based on exclusionary criteria (e.g.,

for random/inconsistent responding and poor understanding of instructions). One other possible

limitation is that using different methods of assessment (e.g., PCL-R vs. PPI, SIRS vs. simulation

instructions) makes it difficult to directly compare results from these two studies. However, these

differences in assessment could also be considered a strength given replication across settings

and assessment modalities, as well as the consistency of the findings.

Despite these limitations, the current investigation does have significant strengths. Use of

two methodologies (analogue simulation and known-groups) allows for both internal control and

generalizability to forensic clinical populations. The use of multiple measures of psychopathy

(both several self-report inventories and the PCL-R) allowed me to examine some of the various

conceptualizations of psychopathy that have been proposed to date, without relying on only one

particular viewpoint. I also used measures with well-established utility in detecting over- and

under-reporting (e.g., the MMPI-2-RF).

Future research in this area should investigate under-reporting in additional samples, such

as forensic samples in which individuals might have significant motivation to present themselves

in an overly positive light (e.g., individuals undergoing pre-employment or custody evaluations).

41

Focus on the “meanness” domain of psychopathy in such studies might, if replicated, provide

further insight as to why they are less able than individuals high on other psychopathic traits and

“normal” individuals to feign psychological adjustment. Because those high on meanness

presented themselves differently than those high on boldness or disinhibition, it is possible that

there may be other differential manifestations between the groups, such as degree or type of

antisocial behaviors. Although psychopathy, in general, did not appear to moderate the ability to

avoid detection on validity scales, it is possible that other variables may have such effects. An

individual‟s level of psychological knowledge and training does not appear to aid in avoiding

detection when feigning (e.g., Marion, Sellbom, & Bagby, 2011). However, other forms of

knowledge (e.g., verbal and quantitative abilities) may increase an individual‟s ability to avoid

detection on validity scales. There is also the possibility that an interaction could be present

between psychopathy and another variable, such that individuals high in psychopathy and high in

intelligence would be better able than individuals low on both of these constructs to avoid

detection on validity scales when feigning. In a broader scope, it may be useful to examine

feigning ability in the context of other areas of study. Some criminological theory rests heavily

on the concept of self-control (e.g., Gottfredson & Hirschi, 1990),. Low self-control, with

resulting high risk for antisocial behavior, has been linked to traits of guiltlessness and

narcissism (Vaugh, DeLisi, Beaver, Wright, & Howard, 2007). Future research should

investigate the possibility that measures of self-control may capture incremental knowledge

above and beyond psychopathy measures in predicting successful feigning ability.

42

References

Arbisi, P.A., & Ben-Porath, Y.S. (1995). An MMPI-2 infrequent response scale for use with

psychopathological populations: The F(p) Scale. Psychological Assessment, 7, 424-431.

Archer, R. P., Buffington-Vollum, J. K., Stredny, R. V., & Handel, R. W. (2006). A survey of

psychological test use patterns among forensic psychologists. Journal of Personality

Assessment, 87, 84-94.

Baer, R. A., & Miller, J. (2002). Underreporting of psychopathology on the MMPI-2: A meta-

analytic review. Psychological Assessment, 14, 16-26.

Bagby, R. M., Marshall, M. B., Bury, A.S., Bacchiochi, J. R., & Miller, L.S. (2006). Assessing

underreporting and overreporting response styles on the MMPI-2. In J. N.Butcher

(Ed.),MMPI–2: A practitioner's guide (pp. 39–69). Washington, DC: American

Psychological Association.

Benning, S. D., Patrick, C. J., Hicks, B. M., Blonigen, D. M., & Krueger, R. F. (2003). Factor

structure of the Psychopathic Personality Inventory: Validity and implications for clinical

assessment. Psychological Assessment, 15, 340-350.

Ben-Porath, Y. S., & Tellegen, A. (2008). Minnesota Multiphasic Personality Inventory–2

Restructured Form: Manual for administration, scoring, and interpretation. Minneapolis:

University of Minnesota Press.

Book, A.S., Holden, R.R., Starzyk, K.B., Wasylkiw, L., & Edwards, M.J. (2006). Psychopathic

traits and experimentally induced deception in self-report assessment. Personality and

Individual Differences, 41, 601-608.

Brinkley, C. A., Diamond, P. M., Magaletta, P. R., & Heigel, C. P. (2008). Cross-validation of

Levenson‟s Psychopathy Scale in a sample of federal female inmates. Assessment, 15,

464-482.

Brinkley, C. A., Schmitt, W. A., Smith, S. S., & Newman, J. P. (2001). Construct validation of a

self-report psychopathy scale: Does Levenson's self-report psychopathy scale measure

the same constructs as Hare's psychopathy checklist-revised? Personality and Individual

Differences, 31(7), 1021-1038.

Butcher, J. N., Graham, J. R., Ben-Porath, Y. S., Tellegen, A., Dahlstrom, W. G., & Kaemmer,

B. (2001). Minnesota Multiphasic Personality Inventory—2: Manual for Administration

and Scoring (2nd ed.). Minneapolis: University of Minnesota Press.

43

Cale, E. M., & Lilienfeld, S. O. (2006). Psychopathy factors and risk for aggressive behavior: A

test of the “threatened egotism” hypothesis. Law and Human Behavior, 30, 51-74.

Camara, W. J., Nathan, J. S., & Puente, A. E. (2000). Psychological test usage: Implications in

professional psychology. Professional Psychology: Research & Practice, 31, 141-154.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:

Erlbaum.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation

analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum Associates.

DeGue, S., & DiLillo, D. (2004). Understanding perpetrators of nonphysical sexual coercion:

Characteristics of those who cross the line. Violence and Victims, 19, 673-688.

Eckman, P. (1985). Telling lies. New York: Norton.

Edens, J. F. (2006). Unresolved controversies concerning psychopathy: Implications for clinical

and forensic decision making. Professional Psychology: Research and Practice, 37, 59-

65.

Edens, J. F., Buffington, J. K., Tomicic, T. L., & Riley, B. D. (2001). Effects of positive

impression management on the psychopathic personality inventory. Law and Human

Behavior, 25, 235–256.

Edens, J. F., Poythress, N. G., & Lilienfeld, S. O. (1999). Identifying inmates at risk for

disciplinary infractions: A comparison of two measures of psychopathy. Behavioral

Sciences and the Law, 17,435–443.

Edens, J. F., Poythress, N. G., & Watkins-Clay, M. M. (2007). Detection of malingering in

psychiatric unit and general population prison inmates: A comparison of the PAI, SIMS,

& SIRS. Journal of Personality Assessment, 88, 33-42.

Flesch, R. (1949). The art of readable writing. New York: Collier Books.

Frick, P. J., & Hare, R. D. (2001). Antisocial Process Screening Device. Toronto, Canada: Multi-

Health Systems.

Gervais, R. O., Ben-Porath, Y. S., Wygant, D. B., & Sellbom, M. (2010). Incremental validity of

the MMPI-2-RF over-reporting scales and RBS in assessing the veracity of memory

complaints. Archives of Clinical Neuropsychology, 25, 49-55.

Hare, R. D. (1985). Comparison of procedures for the assessment of psychopathy. Journal of

Consulting and Clinical Psychology, 53, 7-16.

44

Hare, R. D. (1991). The Hare Psychopathy Checklist—Revised. Toronto, Ontario, Canada: Multi-

Health Systems.

Hare, R. D. (1993). Without conscience: The disturbing world of the psychopaths among us.

New York: Simon & Schuster.

Hare, R. D. (1996). Psychopathy: A clinical construct whose time has come. Criminal Justice

and Behavior, 23, 25-54.

Hare, R. D. (2003). Hare Psychopathy Checklist-revised manual (2nd

ed.). Toronto, Ontario,

Canada; Multi-Health Systems.

Hart, S. D. (1998). The role of psychopathy in assessing risk for violence: Conceptual and

methodological issues. Legal and Criminological Psychology, 3, 121-137.

Hart, S. D., & Hare, R. D. (1997). Psychopathy: Assessment and association with criminal

conduct. In D. M. Stoff, J. Maser, & J. Breiling (Eds.), Handbook of antisocial behavior

(pp. 22-35). New York: Wiley.

Hathaway, S. R., & McKinley, J. C. (1940). A multiphasic personality schedule (Minnesota): I.

Construction of the schedule. Journal of Psychology,10, 249–254.

Hawes, S. W., & Boccaccini, M. T. (2009). Detection of over-reporting of psychopathology on

the Personality Assessment Inventory: A meta-analytic review. Psychological

Assessment, 21, 112-124.

Holden, R. R. (1996). Holden Psychological Screening Inventory manual. North Tonawanda,

NY: Multi-Health Systems.

Holden, R. R., & Evoy, R. A. (2005). Personality inventory faking: A four-dimensional

simulation of dissimulation. Personality and Individual Differences, 39, 1307-1318.

Iverson, G. L., Franzen, M. D., & Hammond, J. A. (1995). Examination of inmates‟ ability to

malinger on the MMPI-2. Psychological Assessment, 7, 118-121.

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and

Psychological Measurement, 20, 141-151.

Kucharski, L.T., Duncan, S., Egan, S.S., & Falkenbach, D.M. (2006). Psychopathy and

malingering of psychiatric disorder in criminal defendants. Behavioral Sciences & the

Law, 24, 633-644.

Jackson, D. N. (1984). Personality Research Form manual (3rd

. ed.). Port Huron, MI: Sigma

Assessment Systems.

45

Levenson, M.R., Kiehl, K.A., & Fitzpatrick, C.M. (1995). Assessing psychopathic attributes in a

noninstitutionalized population. Journal of Personality and Social Psychology, 68(1),

151-158.

Lilienfeld, S. O., & Andrews, B. P. (1996). Development and preliminary validation of a self-

report measure of psychopathic personality traits in noncriminal populations. Journal of

Personality Assessment, 66, 488-524.

Lilienfeld, S. O., & Fowler, K. A. (2006). The self-report assessment of psychopathy: Promises,

problems, and solutions. In C. Patrick (Ed.), Handbook of Psychopathy (pp. 107-132).

New York: Guilford Press.

Lilienfeld, S. O., & Hess, T. H. (2001). Psychopathic personality traits and somatization: Sex

differences and the mediating role of negative emotionality. Journal of Psychopathology

and Behavioral Assessment, 23, 11–24.

Lilienfeld, S. O., & Widows, M. R. (2005). Psychopathic Personality Inventory – Revised

manual. Lutz, FL: Psychological Assessment Resources, Inc.

Lynam, D. R., Whiteside, S., & Jones, S. (1999). Self-reported psychopathy: A validation study.

Journal of Personality Assessment, 73, 110-132.

MacNeil, B.M., & Holden, R.R. (2006). Psychopathy and the detection of faking on self-report

inventories of personality. Personality and Individual Differences, 41, 641-651.

Marion, B. E., Sellbom, M., & Bagby, R. M. (2011). The detection of feigned psychiatric

disorders using the MMPI-2-RF over-reporting validity scales: An analog investigation.

Psychological Injury and Law. Advance online publication. doi: 10.1007/s12207-011-

9097-0.

Millon, T. (1983). Millon Clinical Multiaxial Inventory Manual (3rd

ed.). Minneapolis, MN:

National Computer Systems.

Millon, T. (1987). Manual for the MCMI-II (2nd

ed.). Minneapolis, MN: National Computer

Systems.

Millon, T. (1997). Manual for the MCMI-III (2nd

ed.). Minneapolis, MN: National Computer

Systems.

Morey, L. C. (1991). Personality Assessment Inventory: Professional manual. Tampa, FL:

Psychological Assessment Resources.

Morey, L. C. (1996). An interpretive guide to the Personality Assessment Inventory (PAI).

Tampa, FL: Psychological Assessment Resources.

46

Morey, L. C. (2007). Personality Assessment Inventory (PAI) professional manual (2nd

ed.).

Odessa, FL: Psychological Assessment Resources, Inc.

Mullins-Nelson, J. L., Salekin, R. T., & Leistico, A. R. (2006). Psychopathy, empathy, and

perspective-taking ability in a community sample: Implications for the successful

psychopathy concept. International Journal of Forensic Mental Health, 5, 133-149.

Patrick, C. J. (2008). Triarchic Inventory. Unpublished rating scale.

Patrick, C. J. (2010, March). Operationalizing the Triarchic Model of Psychopathy: Brief scales

for assessment of Boldness, Meanness, and Disinhibition. Symposium presentation at the

2010 American Psychology-Law Soceity Conference. Vancouver, British Columbia,

Canada.

Patrick, C. J., Fowles, D. C., & Krueger, R. F. (2009). Triarchic conceptualization of

psychopathy: Developmental origins of disinhibition, boldness, and meanness.

Development and Psychopathology, 21, 913-938.

Paulhus, D. L. (1998). The Balanced Inventory of Desirable Responding. Toronto, Ontario,

Canada: Multi-Health Systems.

Pearson Education, Inc. (2009). Minnesota Multiphasic Personality Inventory – 2 – Restructured

Form. Retrieved October 5, 2009 from http://psychcorp.pearsonassessments.com/haiweb/

cultures/en-us/productdetail.htm?pid=PAg523.

Poythress, N. G., Edens, J. F., & Lilienfeld, S. O. (1998). Criterion-related validity of the

Psychopathic Personality Inventory in a prison sample. Psychological Assessment, 10,

426–430.

Poythress, N.G., Edens, J.F., & Watkins, M.M. (2001). The relationship between psychopathic

personality features and malingering symptoms of major mental illness. Law and Human

Behavior, 25(6), 567-582.

Rogers, R. (2008). An introduction to response styles. In R. Rogers (Ed.), Clinical assessment of

malingering and deception (2nd ed., pp. 3-13). New York: Guilford.

Rogers, R., Bagby, R. M., & Dickens, S. E. (1992). Structured Interview of Reported Symptoms.

Tampa, FL: Psychological Assessment Resources.

Rogers, R., & Cruise, K. R. (2000). Malingering and deception among psychopaths. In C. B.

Gacano (Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s

guide (pp. 269-284). Mahwah, NJ: Lawrence Erlbaum.

Rogers, R., Gillis, J. R., Dickens, S. E., & Bagby, R. M. (1991). Standardized assessment of

malingering: Validation of the Structured Interview of Reported Symptoms.

Psychological Assessment, 3(1), 89–96.

47

Rogers, R., Sewell, K. W., Morey, L. C., & Ustad, K. (1996). Detection of feigned mental

disorders on the Personality Assessment Inventory: A discriminant analysis. Journal of

Personality Assessment, 67, 629-640.

Rogers, R., Sewell, K. W., Martin, M. A., & Vitacco, M. J. (2003). Detection of feigned mental

disorders: A meta-analysis of the MMPI-2 and malingering. Assessment, 10, 160-177.

Sadeh, N., Verona, E., Javdani, S., & Olson, L. (2009). Examining psychopathic tendencies in

adolescence from the perspective of personality theory. Aggressive Behavior, 35, 399-

407.

Salekin, R. T., Rogers, R., & Sewell, K. W. (1996). A review and meta-analysis of the

Psychopathy Checklist-Revised: Predictive validity of dangerousness. Clinical

Psychology: Science and Practice, 3, 203-215.

Sellbom, M. (2010). Elaborating on the construct validity of the Levenson Self-Report

Psychopathy Scale in incarcerated and non-incarcerated samples. Law and Human

Behavior. Advance online publication. doi: 10.1007/s10979-101-9249-x.

Sellbom, M., & Bagby, R. M. (2008). Response styles on multiscale inventories. In R. Rogers

(Ed.), Clinical assessment of malingering and deception (2nd ed., pp. 182-206). New

York: Guilford.

Sellbom, M., & Bagby, R.M. (2008). Validity of the MMPI-2-RF (Restructured Form) L–r and

K–r scales in detecting underreporting in clinical and nonclinical samples. Psychological

Assessment, 20, 370-376.

Sellbom, M., & Bagby, R. M. (2010). Detection of overreported psychopathology with the

MMPI-2-RF form validity scales. Psychological Assessment, 22, 757-767.

Sellbom, M., Toomey, J. A., Wygant, D. B., Kucharski, L. T., & Duncan, S. (2010). Utility of

the MMPI-2-RF (Restructured Form) Validity Scales in detecting malingering in a

criminal forensic setting: A known-groups design. Psychological Assessment, 22, 22-31.

Smith, G. P., & Burger, G. K. (1997). Detection of malingering: Validation of the Structured

Inventory of Malingered Symptomatology (SIMS). Journal of the American Academy of

Psychiatry and Law, 25, 183-189.

Toomey, J. A., Kucharski, L. T., & Duncan, S. (2009). The utility of the MMPI-2 Malingering

Discriminant Function Index in the detection of malingering: A study of criminal

defendants. Asssessment, 16, 115-121.

Vaughn, M.G., Newhill, C.E., DeLisi, M., Beaver, K.M., & Howard, M.O. (2008). An

investigation of psychopathic features among delinquent girls: Violence, theft, and drug

abuse. Youth Violence and Juvenile Justice, 6, 240-255.

48

Vitacco, M. J., Rogers, R., Gabel, J., & Munizza, J. (2007). An evaluation of malingering screens

with competency to stand trial patients: A known-groups comparison. Law and Human

Behavior, 31, 249-260.

Walters, G. D. (2003). Predicting criminal justice outcomes with the Psychopathy Checklist and

Lifestyle Criminality Screening Form: A meta-analytic comparison. Behavioral Sciences

and the Law, 21, 89-102.

Widiger, T. A. (2006). Psychopathy and DSM-IV psychopathology. In C.J. Patrick (Ed.)

Handbook of Psychopathy (pp. 156-171). New York: The Guilford Press.

Williams, K. M., & Paulhus, D. L. (2004). Factor structure of the Self-Report Psychopathy Scale

(SRP-II) in non-forensic samples. Personality and Individual Differences, 37, 765-778.

Williams, K. M., Paulhus, D. L., & Hare, R. D. (2007). Capturing the four-factor structure of

psychopathy in college students via self-report. Journal of Personality Assessment, 88,

205-219.

Wygant, D. B., Ben-Porath, Y. S., Arbisi, P. A., Berry, D. T. R., Freeman, D. B., Heilbronner, R.

L. (2009). Examination of the MMPI-2 Restructured Form (MMPI-2-RF) validity scales

in civil forensic settings: Findings from simulation and known group samples. Archives

of Clinical Neuropsychology, 24, 671-680.

Wygant, D. B., Sellbom, M., Gervais, R. O., Ben-Porath, Y. S., Stafford, K. P., Freeman, D. B.,

& Heilbronner, R. L. (2010). Further validation of the MMPI-2 and MMPI-2-RF

Response Bias scale: Findings from disability and criminal forensic settings.

Psychological Assessment. Advance online publication. doi: 10.1037/a0020042.

49

Appendices

50

Appendix A

Triarchic Inventory

Directions: This questionnaire contains statements that different people might use to describe

themselves. Most of these statements are followed by four choices: 1-4. The meaning of these

four different choices is given below:

1 = True 2 = Mostly True 3 = Mostly False 4 = False

For each statement, fill in the bubble next to the choice that describes you best. There are no

right or wrong answers; just choose the answer that best describes you.

Remember: Fill only one bubble per item. If you make a mistake cross out the incorrect

answer with an X and fill in the correct option. Answer all of the items. Please work rapidly

and do not spend too much time on any one statement.

1. I‟m optimistic more often than not. O1 O2 O3 O4

2. How other people feel is important to me. O1 O2 O3 O4

3. I often act on immediate needs. O1 O2 O3 O4

4. I have no strong desire to parachute out of an airplane. O1 O2 O3 O4

5. I've often missed things I promised to attend. O1 O2 O3 O4

6. I would enjoy being in a high-speed chase. O1 O2 O3 O4

7. I am well-equipped to deal with stress. O1 O2 O3 O4

8. I don‟t mind if someone I dislike gets hurt. O1 O2 O3 O4

51

9. My impulsive decisions have caused problems with loved ones. O1 O2 O3 O4

10. I get scared easily. O1 O2 O3 O4

11. I sympathize with others‟ problems. O1 O2 O3 O4

12. I have missed work without bothering to call in. O1 O2 O3 O4

13. I'm a born leader. O1 O2 O3 O4

14. I enjoy a good physical fight. O1 O2 O3 O4

15. I jump into things without thinking. O1 O2 O3 O4

16. I have a hard time making things turn out the way I want. O1 O2 O3 O4

17. I return insults. O1 O2 O3 O4

18. I've gotten in trouble because I missed too much school. O1 O2 O3 O4

19. I have a knack for influencing people. O1 O2 O3 O4

20. It doesn‟t bother me to see someone else in pain. O1 O2 O3 O4

21. I have good control over myself. O1 O2 O3 O4

22. I function well in new situations, even when unprepared. O1 O2 O3 O4

23. I enjoy pushing people around sometimes. O1 O2 O3 O4

24. I have taken money from someone's purse or wallet without asking. O1 O2 O3 O4

25. I don't think of myself as talented. O1 O2 O3 O4

26. I taunt people just to stir things up. O1 O2 O3 O4

52

27. People often abuse my trust. O1 O2 O3 O4

28. I'm afraid of far fewer things than most people. O1 O2 O3 O4

29. I don't see any point in worrying if what I do hurts someone else. O1 O2 O3 O4

30. I keep appointments I make. O1 O2 O3 O4

31. I often get bored quickly and lose interest. O1 O2 O3 O4

32. I can get over things that would traumatize others. O1 O2 O3 O4

33. I am sensitive to the feelings of others. O1 O2 O3 O4

34. I have conned people to get money from them. O1 O2 O3 O4

35. It worries me to go into an unfamiliar situation without knowing all the

details.

O1 O2 O3 O4

36. I don't have much sympathy for people. O1 O2 O3 O4

37. I get in trouble for not considering the consequences of my actions. O1 O2 O3 O4

38. I can convince people to do what I want. O1 O2 O3 O4

39. For me, honesty really is the best policy. O1 O2 O3 O4

40. I've injured people to see them in pain. O1 O2 O3 O4

41. I don‟t like to take the lead in groups. O1 O2 O3 O4

42. I sometimes insult people on purpose to get a reaction from them. O1 O2 O3 O4

43. I have taken items from a store without paying for them. O1 O2 O3 O4

53

44. It's easy to embarrass me. O1 O2 O3 O4

45. Things are more fun if a little danger is involved. O1 O2 O3 O4

46. I have a hard time waiting patiently for things I want. O1 O2 O3 O4

47. I stay away from physical danger as much as I can. O1 O2 O3 O4

48. I don't care much if what I do hurts others. O1 O2 O3 O4

49. I have lost a friend because of irresponsible things I've done. O1 O2 O3 O4

50. I don't stack up well against most others. O1 O2 O3 O4

51. Others have told me they are concerned about my lack of self-control. O1 O2 O3 O4

52. It‟s easy for me to relate to other people‟s emotions. O1 O2 O3 O4

53. I have robbed someone. O1 O2 O3 O4

54. I never worry about making a fool of myself with others. O1 O2 O3 O4

55. It doesn‟t bother me when people around me are hurting. O1 O2 O3 O4

56. I have had problems at work because I was irresponsible. O1 O2 O3 O4

57. I‟m not very good at influencing people. O1 O2 O3 O4

58. I have stolen something out of a vehicle. O1 O2 O3 O4

54

Appendix B

Hare‟s Self-Report Psychopathy Scale – 2nd

Edition (SRP-II)

Instructions:

On the following pages you will find a number of statements that have been used by people to

describe their beliefs and behaviors, and the beliefs and behaviors of others. Read each statement

carefully and decide whether you agree or disagree with each statement according to the

following scale:

1 2 3 4 5 6 7

Disagree

Strongly

Disagree

Moderately

Disagree

Slightly

Neutral Agree

Slightly

Agree

Moderately

Agree

Strongly

For example, if you disagree moderately with a statement, write down the number “2” next to it.

If you neither agree nor disagree with the statement, write the number “4” indicating Neutral.

There are no right or wrong answers to any of these questions and none of the questions have any

trick to them. Some of the questions are similar to one another, but judge each one separately. It

does not matter if you have answered a similar question differently--simply indicate how you

would respond to the current statement.

Be sure not to miss any of the questions.

1 2 3 4 5 6 7

Disagree

Strongly

Disagree

Moderately

Disagree

Slightly

Neutral Agree

Slightly

Agree

Moderately

Agree

Strongly

1. I enjoy driving at high speed.

2. I enjoy giving "bossy" people a hard time.

3. I think I could "beat" a lie detector.

4. Sometimes you have to be crafty or sly.

5. It's best to be dominant and assertive because no-one else is going to look out for

you.

6. I worry a lot about possible misfortunes.

55

7. I like to change jobs fairly often.

8. I can be fairly cunning if I have to be.

9. Everybody likes to hear my stories.

10. I am usually very careful about what I say to people.

11. I have often done something dangerous just for the thrill of it.

12. I wish I were more assertive.

13. I expect a great deal from other people.

14. I'm not at all calculating.

15. I think of myself as self-assured and confident.

16. I didn't get into much trouble at school.

17. I get a kick out of "conning" someone.

18. I get in trouble for the same things time after time.

19. I am very good at most things I try to do.

20. I was never in trouble with the police when I was a kid.

21. It's more effective to be straightforward and honest if you want people to do things for you.

22. Being unemployed would depress me.

23. I enjoy taking chances.

24. I wouldn't do anything dangerous just for the thrill of it.

25. I often worry unnecessarily.

26. I insist upon getting the respect that is due me.

27. The best way to get things done is to be forceful and persistent.

28. I got in a lot of trouble at school.

29. Rules are made to be broken.

56

30. I usually feel quite confident when meeting new people.

31. Not hurting others' feelings is important to me.

32. I would be good at a dangerous job because I like making fast decisions.

33. I have used few, if any, hallucinogenic drugs.

34. On average my friends would probably say I am a kind person.

35. I see myself as a good leader.

36. I can read people like a book.

37. I can usually talk my way out of anything.

38. I have used most of the hallucinogenic drugs.

39. I have sometimes broken an appointment because something more interesting came along.

40. I enjoy gambling for large stakes.

41. I have a strong will to power.

42. I would describe myself as a crafty individual.

43. I prefer having many sexual partners rather than just one.

44. I will never be satisfied until I get all that I deserve.

45. One must live only for the present and not worry about the future.

46. If I ruled the world it would be a much better place.

47. Sometimes at night I get so worried about something that my heart pounds and I can't fall asleep.

48. I don't think of myself as tricky or sly.

49. I almost never feel guilty over something I've done.

50. It's sometimes fun to see how far you can push someone before they catch on.

51. People can usually tell if I am lying.

52. I wouldn't describe myself as shy or timid.

57

53. Conning people gives me the "shakes."

54. When I do something wrong, I feel guilty even though nobody else knows it.

55. I always know what I am doing.

56. I find it easy to manipulate people.

57. I'm a soft-hearted person.

58. I enjoy drinking and doing wild things.

59. Ideally people should be undemanding.

60. I am the most important person in this world and nobody else matters.

58

Appendix C

Levenson‟s Self-Report Psychopathy Scale (LSRP)

Instructions: Please respond to each item as honestly as you can. There are no right or

wrong answers, and your answer will be completely anonymous. Using the scale below, place

your response in the space provided next to each item below.

Disagree strongly

1

Disagree somewhat

2

Agree somewhat

3

Agree strongly

4

_____ 1. Success is based on survival of the fittest; I am not concerned about the losers.

_____ 2. For me, what‟s right is whatever I can get away with.

_____ 3. In today‟s world, I feel justified in doing anything I can get away with to succeed.

_____ 4. My main purpose in life is getting as many goodies as I can.

_____ 5. Making a lot of money is my most important goal.

_____ 6. I let others worry about higher values; my main concern is with the bottom line.

_____ 7. People who are stupid enough to get ripped off usually deserve it.

_____ 8. Looking out for myself is my top priority.

_____ 9. I tell other people what they want to hear so that they will do what I want them to do.

_____ 10. I would be upset if my success came at someone else‟s expense.

_____ 11. I often admire a really clever scam.

_____ 12. I make a point of trying not to hurt others in pursuit of my goals.

_____ 13. I enjoy manipulating other people‟s feelings.

_____ 14. I feel bad if my words or actions cause someone else to feel emotional pain.

_____ 15. Even if I were trying very hard to sell something, I wouldn‟t lie about it.

_____ 16. Cheating is not justified because it is unfair to others.

59

_____ 17. I find myself in the same kinds of trouble, time after time.

_____ 18. I am often bored.

_____ 19. I find that I am able to pursue one goal for a long time.

_____ 20. I don‟t plan anything very far in advance.

_____ 21. I quickly lose interest in tasks I start.

_____ 22. Most of my problems are due to the fact that other people just don‟t understand me.

_____ 23. Before I do anything, I carefully consider the possible consequences.

_____ 24. I have been in a lot of shouting matches with other people.

_____ 25. When I get frustrated, I often „let off steam‟ by blowing my top.

_____ 26. Love is overrated.

60

Appendix D

MMPI-2-RF Post-Test Questionnaire

ID #:_______________ Sex:_____________ Age:_______________ Date:_______________

This questionnaire asks questions about how closely you followed the instructions for

completing the MMPI-2-RF test. Please respond HONESTLY regardless of the way you were

asked to answer the questions on the MMPI-2-RF. Moreover, you will receive your participation

points regardless of your answers to this questionnaire, so please be honest.

1. The instructions I received for completing the MMPI-2-RF were:

a. To give an overly positive impression of myself without being detected while taking the

MMPI-2-RF

b. To give an overly negative impression of myself without being detected while taking the

MMPI-2-RF.

c. To complete the MMPI-2-RF in an honest manner.

2. I followed the experimenter‟s instructions on how to take the MMPI-2-RF:

a. All of the time.

b. Most of the time.

c. Half of the time.

d. Less than half of the time.

e. Not at all.

3. I tried to appear as negatively as I could when I took the MMPI-2-RF:

a. All of the time.

b. Most of the time.

c. Half of the time.

d. Less than half of the time.

e. Not at all.

4. I tried to appear as positively as I could when I took the MMPI-2-RF:

a. All of the time.

b. Most of the time.

c. Half of the time.

d. Less than half of the time.

e. Not at all.

5. I tried to be completely honest when I took the MMPI-2-RF:

a. All of the time.

b. Most of the time.

c. Half of the time.

d. Less than half of the time.

e. Not at all.

61

This questionnaire asks questions about how closely you followed the instructions for

completing the MMPI-2-RF test. Please respond HONESTLY regardless of the way you were

asked to answer the questions on the MMPI-2-RF. Moreover, you will receive your participation

points regardless of your answers to this questionnaire, so please be honest.

6. I got tired or bored and answered some of the MMPI-2-RF questions without reading them

carefully:

a. All of the time.

b. Most of the time.

c. Half of the time.

d. Less than half of the time.

e. Not at all.

7. I answered about _________ of the MMPI-2-RF questions without reading them carefully.

a. 0-5

b. 6-15

c. 16-25

d. 26-35

e. 36-45

f. 46-55

g. More than 55

8. I believe that the people not involved with the research (like University Administrators):

a. Will definitely see the results of my test.

b. Will probably see the results of my test.

c. Might see the results of my test.

d. Will probably not see the results of my test.

e. Will definitely not see the results of my test.

9. I believe that there are ways for psychologists to tell if I am answering the MMPI-2-RF

honestly.

a. True

b. False

10. I believe that I was ____________ to fake on the MMPI-2-RF.

a. Definitely Able

b. Probably Able

c. Probably Unable

d. Definitely Unable

e. N/A – I was not asked to fake anything on the MMPI-2-RF

62

Appendix E

Instructions for Under-reporting Condition

Please try to fill out the following questionnaire in order to give a good impression of yourself.

Please pretend that you have been convicted of a crime, and are being evaluated by a

psychologist. You want to make sure that you present yourself in a positive light, so that the

psychologist does not classify you as “difficult,” or “dangerous.” Your answers could decide

whether you are placed in a maximum or minimum security facility. However, keep in mind that

this questionnaire has scales designed to detect dishonest individuals. Thus, try to be as

believable as you can in presenting yourself in a positive manner, without problems, so that you

are not detected by the psychologist.

63

Appendix F

Instructions for Over-reporting Condition

Please try to fill out the following questionnaire in order to give a negative impression of

yourself. Please pretend that you have committed a crime, and are being evaluated by a

psychologist before you go to trial. You decide to fake mental illness, hoping that it will get you

a lighter sentence (or maybe even get you off the hook completely). However, keep in mind that

this questionnaire has scales designed to detect dishonest individuals. Thus, try to be as

believable as you can in presenting yourself negatively, so that you are not detected by the

psychologist.

64

Appendix G

Institutional Review Board Approval Form